Optimizing the Genetic Algorithm to resynthesize iris Patterns from iris Code Template

Human iris is considered to be the most reliable biometric trait. Iris recognition system stores the discriminant and complete information of a particular iris in a compact form known as iris code, because of security reasons. Earlier it was conceived that resynthesizing the original information from the biometric template is not feasible, but later scrutinization proved this notion wrong. Different types of approaches were propounded for resynthesizing biometric data from their corresponding template representations. In this paper, different crossover techniques are used to optimize the genetic algorithm to synthesize iris images from similar iris codes and their performance is evaluated through a commercial matcher by estimating the probability of successfully matching the synthetic iris images with its corresponding real images.